Plenty of methods to synthesize LTO have been reported; however, many of these processes lack of simplicity and are too costly for practical applications.
In this study the hydrothermal method is used to synthesize LTO. It is used three variants of the hydrothermal approach. These three methods are different in their mechanisms which lead to LTO, and as a result the morphologies and the electrochemical performance are different as well.
This chapter is divided in four parts. Part I copes with the hydrothermal synthesis with TiO2 (anatase and rutile) and LiOH in presence of NaOH; this approach is the most conventional approach. In Part II, AHTO is used in presence of LiOH, in this particular synthesis the conditions are softened hence the formation of LTO is easier. In Part III, a new hydrothermal synthesis, using as a precursor TiO2, is explained. The influence in the morphology of the different condition and concentration of the precursors is studied in this Part.
Finally, in Part IV a summary of the result is given.
4-1 Hydrothermal Synthesis: First Method
In general, low temperature LTO phases were prepared by hydrothermal reaction of TiO2 (P25 or rutile) and LiOH with the presence of NaOH in water or alcohol at temperature around 200°C for 20 h, as is illustrated in Figure 3-2. The
33
hydrothermal reaction produced white powder, which was carefully filtered and washed with distilled water to remove an excess of hydroxides and carbonates.
During the treatment with distilled water, the conductivity of the water was controlled and when this property started to drop suddenly to a lower value of 2.5 mS/cm, the process of washing was stopped. Then the powder was dried and finally calcined at 500 oC for 2 hours.
The influence of three variables was investigated: the concentration of LiOH, concentration of NaOH, and the use of solvent (water and alcohol). In Table 4-1 is specified the value of these parameters. It was used two levels for the LiOH, three levels for the NaOH, and two levels for the solvent hence the total number of experiments were 12. The processes are called as follow: for using water Hydrothermal process (HW), and for using alcohol Solvothermal process (SA).
Table 4-1 Variables investigated in the hydrothermal treatment and the stoichiometric ratio Li/Ti used: Fist method
Level 1 Level 2 Level 3
Solvent Water Alcohol 95% -
[LiOH] M 1.50 3.50 -
Stoichiometric ratio Li/Ti 2.5 5.8 -
[NaOH] M 0.00 0.55 1.00
34
For hydrothermal treatment HW, six experiments were carried out and the outcome was the same for all of them, in the Figure 4-1 is presented the XRD pattern for the samples before and after the calcination, the presented pattern are common for all the powders. However, during the preliminary experiments was encountered that with concentration of LiOH less than 1.0 M the powder before the calcination presented some amount of anatase and rutile unreacted, even employing a stoichiometric ration Li/Ti of 4 the latter result persisted. The XRD pattern shown, in Figure 4-1 (before the calcination, lower pattern) presents a small hump at around 2θ = 18°, and strong reflections at 2θ = 43°, 2θ = 63°, and 2θ = 80°, this XRD pattern was encountered before and, actually, has an square structure of lithium titanium oxide; however, as it is explained by Fattakhova et al. [27], the apparent deviation of the experimental diffraction pattern from that expected for spinel structure, namely, diffuse scattering in the range corresponding to 2θ = 18.3° (111) reflection and missing 2θ = 35.5° (311) reflection can be attributed most probably to a disorder in Ti–O framework. In this thesis the latter compound will be called disorder spinel LTO. The synthesized powder was nano crystalline with coherent domain size estimated from SEM micrographs of about 50 nm, as shown the Figure 4-2. After the calcination at 500 oC for 2 hours, yield to a conventional LTO confirmed by XRD measurement, as seen in Figure 4-1.
35
10 20 30 40 50 60 70 80
Intensity
2 Theta (degree)
Before calcination After calcination
Figure 4-1 patterns of the powder before and after the calcination. First method of hydrothermal synthesis with water.
Figure 4-2 SEM micrograph of the powder. First method of hydrothermal synthesis with water.
36
Figure 4-3 XRD patterns for the solvothermal synthesis SA. (a) and (c) LiOH concentration of 1.5 M, with calcination and without calcination respectively. (b) and (d) LiOH concentration of 3.5 M, with calcination and without calcination respectively. A and R indicates the main peaks for antase and rutile, respectively.
Figure 4-3 shows the XRD patterns for solvothermal synthesis SA, in total was carried out 6 experiments. Figure 4-3 (a) and (c) show the XRD patterns for sample synthesized at 1.5 M of LiOH at three levels of NaOH, with calcination and without calcination respectively, the conversion to LTO is almost totally
37
hindered if there is no used of NaOH, but as shown in Figure 4-3 (b) and (c), for 3.5 M of LiOH, the conversion improves with a higher concentration of LiOH.
Figure 4-3 shows that the presence of NaOH is essential in SA for the conversion being completed. As previously explained, in the preliminary experiments was encountered that at lower concentration of 1 M of LiOH the TiO2 is not converted totally, the latter occurs even at larger than 4 of the stoichiometric ratio Li/Ti. Similarly to HW, the SA leads to the disordered spinel LTO, as explained by Fattakhova et al. [27], and the latter can be converted to conventional LTO with thermal treatment (500 oC for 2 hours). According to the SEM micrographs, the morphology is composed of dispersed nano square particles of around 10 nm of size, as displayed in Figure 4-4; hence, SA leads to a smaller particle than HW.
Figure 4-4 SEM micrograph of the powder. Second Method of solvothermal synthesis.
38
The calcination of the powder was carried out at various temperatures between 400 oC to 800oC for 2 hours. For the two powders, i.e. HW’s and SA’s, the tendency is the same, at higher temperature the crystallinity improves, and the peaks in the XRD patterns get shaper; however, between 600 oC and 700 oC, there is a morphology transition where the nano particles melt and join together to form larger particles, as shown in Figure 4-5. As a consequence for the further experiments was chosen an intermediate temperature of 550 oC to keep the same morphology after hydrothermal and obtain good crystallinite.
Figure 4-5 SEM micrographs for the hydrothermal and solvothermal synthesis for two calcination temperatures, 600 oC and 700 oC for 2 hours. The melting and growing of the particles could be observed at 700 oC.
39
The hydrothermal treatment leads to the same products regardless whether rutile or anatase powders were used as starting source of TiO2. Hence the reaction has to proceed most likely via dissolution – precipitation mechanism.
4-2 Hydrothermal Synthesis: Second Method
In previous reports, the AHTO was synthesized first then the hydrothermal synthesis was performed, as example we can cite Kabac et al. [39], Qiu et al. [40], and Tang et al. [29]. We follow a similar approach , in Figure 3-3 shows the schematic diagram of the process.
Mono dispersed AHTO microspheres were prepared by controlled hydrolysis of titanium ethoxide, this precursor was used as received from Wu et al [38]. In a typical process of fabricating LTO, 412 mg of AHTO were dispersed in 20mL, 0.4M LiOH solution (stoichiometric ratio Li/Ti around 2.5). After stirred for 10 min, the suspension was transferred into a 40mL Teflon-lined stainless steel autoclave and heated at 185 oC for 8 h. The hydrothermal reaction produced white powder, which was carefully washed with distilled water to remove an excess of hydroxides. Then the powder was separated from the washing solution by filtration, dried in an oven, and finally calcined at 550 oC for 24h. For AHTO presents an amorphous phase according to the XRD pattern and after the hydrothermal treatment the disordered spinel LTO is synthesized, further calcination leads to conventional LTO, as shown the XRD patterns of the powders presented in Figure 4-6. It is believed that the uptaking of Li follows the mechanism analogous to the Kirkendall effect as is explained by Wang et al. [41].
The morphology of the AHTO and LTO synthesized by this method are shown in
40
the micrographs of Figure 4-7(a, b). According to the XRD pattern, following the Scherrer formula, delivered approximately 20 nm for particle size.
Figure 4-6 XRD pattern of the powder synthesized by hydrothermal synthesis of AHTO.
Figure 4-7 SEM micrographs (a) AHTO; (b) powder synthesized by hydrothermal treatment after calcination.
41 described below leads to different morphologies: nano flakes or quasi-square nano particles, those forming aggregated secondary particles or dispersed alone, by controlling the parameters in the hydrothermal synthesis such as ratio Li/Ti, temperature, concentration of H2O2, and pH.
Figure 3-4 presents the process diagram flow of the hydrothermal process to produce LTO using titanium peroxide. TiO2 is stirred in basic hydrogen peroxide solution, and then this mixture is placed into a Teflon-line for the hydrothermal synthesis. After the reaction, the precipitate is filtrated and treated with deionized water, this step of procedure is repeated until the pH and conductivity begins to drop. At this point, the precipitate is dried and finally calcined in a furnace.
During the preliminary experiments, after the hydrothermal reaction, it was noticed that the treatment with deionized water is an essential part to obtain pure LTO, to elucidate this phenomenon the next experiment was designed measuring the pH and the conductivity during every treatment with deionized water. The hydrothermal treatment of solution 0.2 M titanium peroxide, 0.4 M LiOH, and 0.89 M H2O2, was carried out at 150 °C for 12 hours (the
42
stoichiometric ratio Li/Ti was 2.5). For about one gram of synthesized powder the measurement of pH and conductivity through the step of treatment with deionized water are shown in the Figure 4-8. The conductivity and pH through the treatment with deionized water shows an impartially invariant value till certain point where suddenly begin to drop reaching values around 0.45 mS/cm and 10.5, respectively, (particular values for this experiment). The XRD patterns of calcined powder, Figure 4-9(a) shows that for the step of process before both conductivity and pH dropped, there are clear reflections at around 2θ = 20.3°, 2θ
= 63.5° and 2θ =66.8°, which can be indexed to the Li2TiO3 (JCPDS 00-033-0831) which is another phase of lithium titanium oxide. The latter compound was encounter before [42] when there is an excess of lithium in the powder, and the conclusions are that this compound diminish the capacity. Hence, for these samples there is a mixture of LTO and Li2TiO3 rich in lithium. Upon the steps of treatment with deionized water, the concentration of Li2TiO3 decreases until a pure phase of pure Li4Ti5O12 can be obtained, for further treatments a deficiency of lithium is encountered, and consequently the phases of anatase and rutile of TiO2 emerge as is presented in the uppermost curve in Figure 4-9(a). More clearly, the shift to the left of the reflections at around 2θ = 63.5° is noticed, at larger angle where the peak shifts due to the lattice parameter changes are more distinguishable. Specifically, the reflection of Li2TiO3 at 2θ = 63.5° (-2 0 6) fades through the washing steps upon is only noticed the reflection of Li4Ti5O12 at 2θ = 62.83° (4 4 0) as is presented in Figure 4-9(b). As the SEM micrographs revealed, the morphology neither changes over the steps of treatment with deionized water
43
nor changes after the calcination at 550 °C for 4 h. Figure 4-10 shows the morphology consisting in nano flakes mixed with some square particles for the sample taken right after the hydrothermal synthesis, at the 13th step of washing before the calcination, and at the 13th step of washing after the calcination.
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Figure 4-8 Plot of the conductivity and pH of the filtered water, and dilution factor of the sample vs. the step of process for 1 gram of precipitate.
44
Figure 4-9 Powder products obtained at 150 °C for 12 h hydrothermal synthesis at increasing step of treatment with deionized water: (a) XRD patterns of the powder products after calcination at 550 °C for 4 h, and (b) closer look of the XRD patterns. A, R, and L stand for anatase, rutile, and Li2TiO3, respectively.
45
Figure 4-10 SEM micrographs of the powder products obtained at 150 °C for 12 h hydrothermal synthesis, the samples were taken (a) right after the hydrothermal, (b) at the step 13th before the calcination, and (c) at the step 13th after the calcination.
4-3.1 The Formation of Titanium Peroxide and the Hydrothermal Process It is worth to mention that employing antase in the mixture leads to its dissolution, this occurs in the basic solution and presence of hydrogen peroxide at ambient temperature. As mentioned above, there are many ways to obtain the aqueous soluble titanium peroxide, and most of them imply mixing organotitanium or soluble titanium with H2O2. The equations eq. 2-4 and eq. 2.5 describe the global reaction which produces a mononuclear complex. How the mechanism of the formation of the titanium peroxide proceeds is still in question.
Differently from the other approaches, in this study, the formation of clear solution of titanium peroxide was attained via the disolution of anatase in basic solution of H2O2 at room temperature (25 °C). The whole preparation procedure does not involve any use of either organic compound or chlorine-containing
46
hazards. A clear solution was obtained with different sources of hydroxyl (NaOH, LiOH, KOH, and NH3); e.g., 0.1 M of titanium (equivalent to anatase) was dissolved in a solution of 0.89 M H2O2 and 0.4 M NaOH, LiOH, KOH, or NH3 at room temperature, a clear solution was attained as shown in the Figure 4-11. The elapsed time of dissolution for the hydroxyls was similar, but for the NH3 solution longer time was required. In order to detect the soluble aqueous titanium in the form of the titanium peroxide, the pH was decreased by adding acid, upon this step is obtained a deep red color of the mono nuclear peroxo titanatic acid, TiO2(OH)+. After diluting until 1 mM of titanium and adjusting the concentration of hydronium to 1 M with sulfuric acid, the solution was scanning with the colorimetric measurement, delivering a maximum absorbance at around 410 nm [33] wavelength, which confirms the presence of the peroxo titanate acid.
Certainly, the dissolution of the anatase is related with the presence of hydroxyl. At neutral pH, employing 0.1 M of anatase, very little amount of the nano crystal where dissolve, this was accompanying with the decrease of pH from neutral to 4.5, the colorimetric measurement of the filtrated acid liquid indicated the present of peroxo titanate acid. In other experiments, it is noticed that for solutions of higher pH (larger amount of OH-) the amount of dissolved anatase is also larger, while the declining of the pH takes place at the same time.
The latter phenomenon indicates the presence of some acids.
47 Figure 4-11 Clear solution of titanium peroxide
During the experiments, a solution with 0.89 M of H2O2, 0.2 M hydroxyl was able to dissolve 0.2 M of anatase, and 0.4 M hydroxyl dissolve 0.4 M of anatase, the pH of both systems drop to 8.68 and 9.73, respectively. The remaining of pH in the basic scale indicates the presence of the dinuclear titanium peroxides which co-exist at this pH [32]. The basic titanium peroxides solutions prepared were unstable and eventually precipitated, with increasing the pH at approximately to the starting value. Also in another experiment, P25 was mixed with basic solution of H2O2 and this did not dissolve completely; it was obtained a yellow suspension with many particles remaining without dissolving.
This could be due to the presence of rutile in the mixture of TiO2 in P25.
Moreover, rutile (PT-501R, particle size between 100 and 300 nm) was mixed
48
with basic hydrogen peroxide solution, showing not perceivable changing;
however, the acidified filtrated liquid during the colorimetric measurement indicated that there was titanium peroxide in solution. From the latter experiment, it can be inferred that rutile is less soluble than anatase at room temperature (25
°C), and this solubility is limited at certain point. On the basis of our experimental results, we speculate that the overall reaction for the dissolution of titanium dioxide and formation of titanium peroxide can be written as follow:
TiO2(s) + H2O2 + OH- ====> Ti(O2)(OH)3- (4-1) In which the anatase phase is easier to dissolve than rutile. However, the presence of the mononuclear aquo-titanium acid is doubtful, this certainly does not occur in basic pH rather the equilibrium of dinuclear complexes are present, as Schwarzenbach et al. [32] explained. The mechanism of the dissolution of anatase remains unknown yet, and more research has to be done to make more conclusions.
To understand further the hydrothermal treatment, in the preliminary experiments, different morphologies of TiO2 were used, for example, a set of experiments were carried out as follow: first, 0.1 M titanium equivalent to pure rutile or anatase, or a mixture of them was mixed with 0.4 M LiOH and 0.89 M H2O2 (the stoiquiometric ratio Li/Ti was 5), and then the hydrothermal treatment was carried out for 12 hours at 130 °C, 150 °C, and 180 °C, separately, finally, calcination at 550 °C for 4 h was carried out. Figure 4-12(a) shows the XRD pattern of the synthesized powder before the calcination for the synthesis with rutile, anatase, and P25 at lower temperature lower 130 oC. For the synthesis with
49
rutile the conversion is function of the temperature as shown in the Figure 4-12(b). For the latter experiments with rutile, the patters show sharp reflections of rutile; conversely, for the synthesis with anatase and P25 the conversion was completed at all tested temperatures. More importantly, the SEM micrographs reveal that the morphology does not either change starting with rutile or with anatase, or with the mixture of them. In Figure 4-13 shows the SEM images of the synthesized powder at 130 °C before the calcination from rutile, anatase, and P25, which comprises nanoflakes agglomerated in a spherical secondary particle, this morphology coincides identically with the one starting from anatase. The micrographs for the synthesis with rutile reveal that there are some particles of unreacted rutile which contrast with the synthesis with anatase; this observation is consistent with the XRD measurement. Consistently, phenomenon occurs when the starting precursor is P25 and treated at 130 oC, that is, a complete conversion is achieved and the spherical agglomeration of nano flakes is observed in SEM. On the basis of the latter experiments, it can be concluded that the mechanism goes through the same path leading to the same morphology and products regardless of whether rutile or anatase powders were used, or regardless the particle size of the rutile, thus the reaction has to proceeds most likely via first the dissolution of the TiO2 with the reaction eq. 4-1 and then the dissolution-precipitation mechanism of lithium titanium oxide.
50
51
Figure 4-13 Hydrothermal processes were carried out with the concentrations 0.4 M LiOH, 0.1 M TiO2 (rutile, anatase, and P25) and, 0.89 M H2O2 at 130 °C for 12 h.
52
4-3.2 Morphology Control by Hydrothermal Synthesis
For the hydrothermal synthesis via peroxo titanate, the morphology was greatly influence by the temperature, stoichiometric ratio Li/Ti, concentration of H2O2, and the concentration of OH-. For the following synthesis, a clear solution of titanium peroxide in basic solution from anatase was prepared previously.
Figure 4-14 XRD patterns of the powders obtained by hydrothermal treatment of 0.1 M titanium peroxide, 0.4 M LiOH, and 0.89 M H2O2 at 130 °C, 150 °C, and 180 °C for 12 h, before the calcination.
53
Figure 4-15 (a) and (b) SEM micrographs of the powder before the calcination;
the hydrothermal treatment was carried out with the concentrations 0.4 M LiOH, 0.1 M titanium peroxide and, 0.89 M H2O2, at 130 °C for 12 h. (c) TEM micrograph of the powder.
First, to understand the influence of the temperature, the hydrothermal treatment of the solution 0.1 M titanium peroxide (from antase) and 0.4 LiOH was carried out at 130 °C, 150 °C, and 180 °C for 12 h, the stoichiometric ratio of Li/Ti and
54
the concentration of H2O2 were held at 5 and 0.89 M, respectively. Figure 4-14 shows the XRD patterns of the powders before the calcination for the three temperatures of the hydrothermal treatment. At 130 °C all the reflections of the XRD pattern can be ascribed to the layered lithium titanate (LLT) (JCPDS 47-0123) which has a C-base-centered orthorhombic lattice [43], this crystal has a composition of (Li1.81H0.19)Ti2O5 2.2H2O; however, some of the lithium was interchange by hydrogen during the treatment with deionized water to form H2Ti2O5, this interchange is quintessential to obtain pure LTO after the
the concentration of H2O2 were held at 5 and 0.89 M, respectively. Figure 4-14 shows the XRD patterns of the powders before the calcination for the three temperatures of the hydrothermal treatment. At 130 °C all the reflections of the XRD pattern can be ascribed to the layered lithium titanate (LLT) (JCPDS 47-0123) which has a C-base-centered orthorhombic lattice [43], this crystal has a composition of (Li1.81H0.19)Ti2O5 2.2H2O; however, some of the lithium was interchange by hydrogen during the treatment with deionized water to form H2Ti2O5, this interchange is quintessential to obtain pure LTO after the